Analysis Of Stock Market Based On Cellular Automata

As a typical complex adaptive system, the internal financial market has multiple levels and the characteristics of participation elements and complex nature, variable relationship etc. The traditional research methods focus on the overall grasp of finance, ignore the interaction from individual to individual. This makes the traditional financial theory by more and more questions. People need a new nonlinear method to study the financial market. The complexity of the last century developed scientific success in dealing with nonlinear problem,which has led many scholars began to study with the aid of complexity tools to build the model to study the stock market. As a powerful tool for exploring complexity and complex system, cellular automaton are being widely used in the research of the stock market.In existing studies, there have been a number of cellular automaton model was put forward, although to a certain extent, can simulate and explains the macroscopic changes in stock prices, but some important influence factors are neglected. In this paper, a new cellular automaton model is established for the improvement of these problems,The model compares with the last model has the following features:(1)There is different with the previous model does not consider investors’ funds.This paper will be the overall amount of capital investors in the form of a normal distribution random variables into consideration, more in line with the stock market situation; (2)We do a more detailed description of the moving average trading strategy, considering the influence of the average slope size on the investors’ decision making, which makes the model more realistic; (3)It could divide investors into fundamental investors and technical investors, and introduces the proportional coefficient, studies the evolution of the stock price and yield under different proportion of investors.On the basis of the classical model we introduce genetic algorithm was improved, with the help of biological evolution mechanism to design the meta intercellular evolution rules, and giving cell self-learning ability, rather than in the previous model simply imitate each other, making the model has more practical significance.The improved model has the following features:(1)The investors are no longer a single use a kind of trading strategies, who can use at the same time two trading strategy, strategy comes with its own weight. Introducing position weight at the same time, attribute to investors made a more detailed depiction; (2)The genetic algorithm into the cellular automaton model, a redesign of cellular evolution rules.that can give cell self-learning ability, and increase the intelligent degree of the model; (3)We study the evolution of the stock price, the yield and the weight coefficient separately under the different learning ability of the investor.Finally, it will lead a conclusion of this papar that by using the MATLAB simulation of the model, combining with the two models:(1)Technical investors excessive speculation can cause sharp fluctuations in the market; (2)The average yield of technical investors more than the fundamentals of the investors, the investors have reference value: (3)Learning mechanism is better than imitation mechanism.But investors’blindly learning can also lead to increased market volatility, in moderate learning and their own independent thinking, develop a reasonable trading strategy with its own characteristics, which can maximize profits,In a certain extent can also maintain market stability; (4)The introduction of genetic algorithm allows investors to develop more flexible trading strategies to improve the profits of investors,When the market volatility, improve the weight of the technical side of the strategy, reduce the position, the market stable operation, you can provide the weight of the basic strategy, improve the position.